EP3340104A1 - Verfahren zur alarmerzeugung in einem videoüberwachungssystem - Google Patents
Verfahren zur alarmerzeugung in einem videoüberwachungssystem Download PDFInfo
- Publication number
- EP3340104A1 EP3340104A1 EP16205866.3A EP16205866A EP3340104A1 EP 3340104 A1 EP3340104 A1 EP 3340104A1 EP 16205866 A EP16205866 A EP 16205866A EP 3340104 A1 EP3340104 A1 EP 3340104A1
- Authority
- EP
- European Patent Office
- Prior art keywords
- individual
- targeted
- action
- alert
- actions
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 64
- 230000009471 action Effects 0.000 claims abstract description 108
- 230000004044 response Effects 0.000 claims abstract description 3
- 238000012545 processing Methods 0.000 claims description 22
- 238000013459 approach Methods 0.000 claims description 5
- 238000011156 evaluation Methods 0.000 claims description 5
- 230000002123 temporal effect Effects 0.000 claims description 5
- 238000004590 computer program Methods 0.000 claims description 4
- 230000002596 correlated effect Effects 0.000 claims description 3
- 230000008901 benefit Effects 0.000 description 7
- 239000013598 vector Substances 0.000 description 7
- 238000001514 detection method Methods 0.000 description 6
- 238000004891 communication Methods 0.000 description 4
- 238000012552 review Methods 0.000 description 4
- 241000282326 Felis catus Species 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 238000010420 art technique Methods 0.000 description 2
- 230000015572 biosynthetic process Effects 0.000 description 2
- 238000012790 confirmation Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 239000003550 marker Substances 0.000 description 2
- 241001465754 Metazoa Species 0.000 description 1
- 230000016571 aggressive behavior Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000003139 buffering effect Effects 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000003203 everyday effect Effects 0.000 description 1
- 230000000977 initiatory effect Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/20—Movements or behaviour, e.g. gesture recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
- G06T7/246—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/41—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B13/00—Burglar, theft or intruder alarms
- G08B13/18—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength
- G08B13/189—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems
- G08B13/194—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems
- G08B13/196—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems using television cameras
- G08B13/19602—Image analysis to detect motion of the intruder, e.g. by frame subtraction
- G08B13/19608—Tracking movement of a target, e.g. by detecting an object predefined as a target, using target direction and or velocity to predict its new position
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B13/00—Burglar, theft or intruder alarms
- G08B13/18—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength
- G08B13/189—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems
- G08B13/194—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems
- G08B13/196—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems using television cameras
- G08B13/19602—Image analysis to detect motion of the intruder, e.g. by frame subtraction
- G08B13/19613—Recognition of a predetermined image pattern or behaviour pattern indicating theft or intrusion
- G08B13/19615—Recognition of a predetermined image pattern or behaviour pattern indicating theft or intrusion wherein said pattern is defined by the user
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/18—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
- H04N7/183—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast for receiving images from a single remote source
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
Definitions
- the present invention relates to a method for tracking objects, and in particular to a method for generating alerts in a video surveillance system.
- action recognition is an area of development.
- the purpose of action recognition is to enable automatic detection of actions performed by objects in a surveilled scene, generally human actions.
- An example could be that while a perimeter sensor could generate an alert when an individual or object is entering a particular area, action recognition may generate an alert when the same individual assumes a hostile pose.
- the development is catalyzed by improved image processing techniques, deep learning etc, enabling more advanced processing in shorter times.
- the information typically includes a number of live video streams. More often than not there is a recording function enabling review of the information afterwards, but there is a benefit in being able to react momentarily to unwanted actions occuring in one of the surveilled scenes rather than only being able to analyze in retrospect. In theory it would be possible to track and analyze everything being caught by the video camera. In reality both the human factor and current limitations in processing performance make such an approach, i.e. live evaluation of all information caught by a number of video cameras, unrealistic.
- the present invention aims at providing an improved operator support, in partilcar for complex surveillance situations.
- a surveillance system motion video camera
- detecting a targeted individual in a video stream selecting the targeted individual, tracking the targeted individual, classifying actions of the detected individual over a plurality of image frames in the video stream in response to the identification of the detected object as being a targeted person, and generating an alert signal if the classified action of the object is classified as a predefined alert-generating action.
- Identification of an individual may be based on any existing technique for the technical field. Often an output includes an appearance model (such as appearance vector or feature vector) that may be compared to a stored appearance model.
- an appearance model such as appearance vector or feature vector
- the present invention addresses the need for optimizing tracking processes and in particular the generation of alerts. From a surveillance point of view there is a benefit in being able to backtrack the actions of individuals in order to sort out a particular chain of events, and as such video surveillance has been a highly useful tool throughout its application. Video surveillance has also been effective in focussing on particular individuals so as to e.g. catch someone in the act of performing a prohibited action.
- the surveillance system may be a motion video camera.
- the action recognition is performed on a subset of the image frame, e.g. a crop including the targeted individual but not the entire image frame.
- a subset of the image frame e.g. a crop including the targeted individual but not the entire image frame.
- the computational load is reduced significantly, in particular for a view containing a large number of individuals. This of course implies that some individuals in the scene will be disregarded (only the individuals having been selected and tracked will be actively surveilled).
- Comparing a surveilled individual with the contents of a database is a straightforward way of initiating a tracking sequence, irrespective if the identification and/or the comparison is performed manually or automatically.
- the system may output a invidualized feature vector or appearance vector for later reference.
- the identification of a target person is performed by a recognition algorithm, whereby the identification could be automized to the benefit of an operator.
- the operator may or may not be alerted that an individual has been identified and that tracking has been initiated.
- the operator may be prompted whether the tracking should be discontinued or not.
- the database with which comparison is made, manually or automatically, may comprise previously identified individuals, and in the present context identified implies identified by name, or merely identified in the sense that their identity remains unknown but that they have been the subject of some previous control or surveillance.
- the database may comprise images of the identified individuals and/or feature vectors representing the identified individuals.
- the database may comprise sets of identified attributes relevant for a particular surveillance situation or location. Examples of attributes could be 'carries a bag', 'eyes not visible'/'face masked', etc. i.e. attributes that may not be suspicous in an everyday situation, but depending on context it may be worth to note.
- the predefined alert generating action may be selected in a database comprising a set of "prohibited" actions.
- This embodiment has several levels of complexity. In the first level there may only be a certain number of actions that may be identified, and among these a portion are considered prohibited. Furthermore, for a specific surveillance situation one set of actions would be considered prohibited while in another another set of actions would be considered prohibited. These sets would be different in a warehouse compared to in a sports arena or a railway station.
- a preset set of actions may be alert generating based on the surveillance situation ("in this area the following actions will be prohibited”), yet in other or related embodiments the set may be decided based on a particular individual ("for this person the following action would be prohibited and alert generating") e.g. for a known shoplifter. This will be further addressed in the detailed description.
- the selection of alert generating actions may be correlated to the individual and/or to a particular area.
- the selection of alert generating actions may be correlated to a particular individual. Examples include that the operator or the system recognizes an individual having performed prohibited actions before, and in such a case the individual could be tracked for the event that the prohibited action is repeated. This embodiment could be performed without interaction of the operator, at least until an alert has been generated.
- the selection of alert generating actions may be associated with a particular area, and as an example selections could include that in some areas the most common prohibited actions may be running and looting, while in others it could be "showing signs of aggression", "formation of a group of individuals” etc. This is yet another way of reducing the computational load in connection with targeted action recognition.
- the selection of alert generating actions is prompted upon selection of a targeted individual.
- selection window is presented as an operator identifies a person, e.g. by clicking on that person.
- the operator may select a number of actions that would be considered prohibited for that particular person.
- the available selections could correspond to all availabe actions that may be recognized, or the available selections could also be limited to a subset of actions relevant for the particular individual, e.g. prohibited actions that the individual is known to have performed before.
- a selection pane may be arranged to appear as the individual is identified and the operator may dismiss the pane or make a selection to be activated.
- a selection pane is merely an example among many possible ways of presenting a query.
- the action recognition is at least partly performed by an image processing algorithm in the camera.
- the action recognition may also be performed in its entirety by an image processing algorithm in the camera.
- an action candidate recognizer is configured to apply a first action recognition algorithm to the object image frame and thereby detecting presence of an action candidate; a video extractor configured to produce action image frames of an action video sequence by extracting video data pertaining to a plurality of image frames from the video sequence, wherein one or more of the plurality of image frames from which the video data is extracted comprises the object of interest; and a network interface configured to transfer the action video sequence to the server.
- the server comprises an action verifier configured to apply a second action recognition algorithm to the action video sequence and thereby verify or reject that the action candidate is an action.
- the action classification may be performed using a camera-based image analytics (the image processing algorithm), while in other or related embodiments the action classfication is supported or performed by a remote device as discussed above.
- the method may further comprise a a handover from a first device to a second device of the surveillance system, including communication of a an appearance vector to a network of video cameras.
- an appearance vector of the targeted object is recorded to a database, either as a new post or as an update.
- the action recognition may be based on a spatial/contextual recognition approach, or a temporal recognition approach.
- a system comprising a digital network camera, having an image processing circuitry, an object identifier, an object tracker, an alert generator, wherein the latter may form part of the digital network camera or be comprised in a server-based solution (a server or a network of servers), and the system further comprising a user interface, wherein said system is configured to perform the method according to one or any embodiments disclosed in the present application.
- a computer program product comprising a computer-readable storage medium with instructions adapted to carry out the method according to the description above or below when executed by a device having processing capability.
- a single operator may be responsible for monitoring several camera views simultaneosly.
- surveillance cameras may cover the entrance, the checkout line and a number of positions inside the store. If a suspicious individual is targeted this will imply that the awareness of other views is lowered, since the operator will have an increased awareness in relation to the targeted individual.
- the operator may make an active choice when a targeted person is selected for tracking, i.e., identified as an object to be tracked.
- a targeted person i.e., identified as an object to be tracked.
- the attention of the operator may be turned elsewhere, and the system will generate an alert signal if one of the prohibited actions in conducted, whereby the operator may be notified for full review of the situation.
- Fig. 1 illustrates the inventive method according to a first embodiment 100 thereof.
- the method 100 is initatied by a step of detecting 102 a targeted individual in a scene.
- the individual is yet to be targeted, but for the sake of simplicity the term "targeted individual” is used rather than "an individual which possibly may be targeted at a later stage".
- this step is often preceded by an automated step performed by an object detection algorithm or motion detection algorithm that recognizes that an individual has entered the imaged scene, and that the individual is located in a particular area, often resulting in that the individual (or other moving object) is high-lighted by a rectangle.
- the operator will select the targeted invidual. If the indivual is already marked with a rectangle or in any other way, the selection is likely to be conducted by using a marker to click, in the view, in the near vicinity of the individual in question.
- the tracking 106 of the targeted individual may be initiated automatically or upon user request, by means of communication via any suitable interface. Furthermore, a set of actions that may be conducted by the target individual are classified 108 as prohibited (as defined in the context of the present description). This will be discussed in more detail referring to other embodiments, but it may be stated already that the classification of actions to "prohibited” may be performed in many different ways, both manual and automized.
- the system will generate an alert 112 as soon as it is detected that the targeted individual performs an action 110 included in the set of prohibited actions.
- the alert may be used for several purposes, e.g. it could merely generate a marker for later review, or it could generate an immediate notification to an operator as soon as the action is detected. If an action not included in the list of prohibited actions is identified the system will merely return to tracking and action recognition, without generating any alert.
- An individual may be detected in many different ways, and the step of selecting the individual may be closely related to the detection.
- the individual is detected and targeted merely by being selected by the operator, there is no further logic behind it.
- the operator could be supported by a database 116 containing individuals of particular interest.
- the database may or may not be integrated with the system performing the method.
- the operator may also be supported by a recognition system 114 presenting a likely identity of the individual. If it is a fact that a particular individual will always be selected, e.g. a known shoplifter entering a store, the entire step of detecting and selecting may be performed more or less automatically, with or without sending an alert to the operator that tracking as been initiated.
- the individual is not identified based on personal appearance but rather on attributes 118.
- the attributes may be physical, such as if the person is carrying a bag, pushes a trolley, carries an unappropriate tool etc. To sum up, this portion of the method may be very simple, yet it may also be elaborate and only limited by available technology.
- the prohibited actions are preset, meaning that regardless of the particular targeted individual the same set of actions will be considered prohibited. This may in turn have a further alternative in that the selection depends on a particular area rather than on a particular individual.
- the set of prohibited actions is defined by the operator, and in still other or related embodiments the set of prohibited actions is fully or partly defined by the identified individual. The latter does not imply that the identified individual performs the selection but that for a particual individual a tailormade set of actions may be fully or partly preset. No matter how the set of prohibited actions is deduced, an operator may - in one or more embodiments - be able to finetune the set by adding or removing actions.
- the set of prohibited actions may be associated with a location, meaning that a particular action is prohibited in a particular area only, such that an alert is generated only if the action is performed within that particular area.
- the actions in question could also be related to a class of objects, e.g. it could differentiate between a child and an adult, an employee and a non-employee, and the set of alert-generating actions could be adjusted accordingly.
- the actions in question could be related to an individual object, e.g. this person should not be allowed to carry a bag in this area, etc.
- the step of tracking the individual may be conducted in any of the existing ways. If a single camera is used and the individual never leaves the scene the tracking is quite straightforward, and could be performed according to one of the techniques "tracking by detection” 120 or “motion based tracking” 122, or by another or combined present or novel technique ensuring that the same individual is tracked over a single or several scenes, such as tracking by re-identification.
- the present invention has distinct advantages in systems where several surveillance cameras are involved, since that is a typical situation where the operator may benefit from extra support.
- an appearance model may be made available for all cameras within the system, either by being stored 124 at a location available to all cameras (or at least the relevant ones), or broadcasted 126 to other cameras of the system.
- Action recognition is an enabler for the present invention rather than a main issue, but still some alternatives may be mentioned. Some have already been mentioned in the background section of the present application, and the conclusion is that any method for action recognition may be used. There are however two features that may be relevant for the field of video surveillance. One is that the action recogition should be performed on the fly, since an immediate alert is appreciated, and the other is that there may be limitations to the processing performance available onboard the camera. For straightforward action recognition techniques, such as techniques based on spatial or contextual information 128, these features rarely result in any problems. However, for state of the art techniques where a large amount of data have to be processed it may generate problems, e.g. that the processing performance is inadequate. Examples of the latter are techniques where a temporally resolved sequence of images is used 130.
- Contextual and/or spatial action recognition algorithms are for example described in CN102855462 , in CN103106394 , and by Karen Simonyan and Andrew Zisserman in "Two-Stream Convolutional Networks for Action Recognition in Videos”; arXiv:1406.2199.
- the action recognition is performed as a two-stage process 132.
- an initial recognitions is made, suggestively based on an action recognition technique of low complexity, such as a technique only evaluating a current pose of the targeted individual.
- a confirmation is made using a more complex technique, such as a technique using temporally resolved information. This enables a quick 'on-the-fly' alert followed by a confirmation.
- the information e.g. image frames or crops of image frames, necessary for performing the action recognition may be transferred to a server 134 having superior computational power compared to the camera.
- the alert generation could be performed in several different ways, including techniques used for alert generation today as well as novel ways of generating alerts, such as sending a signal to a VMS alerting the operator, sounding an alarm, performing any preset action.
- Fig. 5 is a schematic view of a system 140 configured for performing an inventive method according to one or any of its embodiments.
- the system comprises one or more network cameras, which is more or less all that is required.
- the digital network camera 142 is arranged to capture a video sequence depicting a scene.
- the digital network camera 142 comprises a housing 144, a lens 146 and circuitry 148.
- the digital network camera 142 is arranged to capture and process, and in some embodiments also store a video sequence.
- the circuitry 148 comprises an image sensor 150, an image processing unit 152, an object identifier 154, an action candidate recognizer 156 and a network interface 158.
- the circuitry 148 may further comprise one or more of a central processing unit, CPU, 160, a digital data storage medium (memory) 162 and an encoding unit 164.
- any one of the image processing unit 152, the object identifier 154, the action candidate recognizer 156, and the encoding unit 164 may be implemented as a dedicated hardware circuit and/or software module.
- the software may be run on the CPU 160.
- the CPU 160 may be any suitable CPU for performing digital data processing.
- any dedicated hardware circuit may in part comprise software portions being run on a dedicated processor or on the CPU 160.
- the digital network camera comprises a tracker (not shown) which is a function performed by the already described components.
- the memory 162 may be any kind of volatile or non-volatile memory. Further, the memory 162 may comprise a plurality of memory units. At least one of the plurality of memory units may be used as a buffer memory for buffering data while processing e.g. content of the video sequence.
- the digital network camera 142 is arranged to be connected to a digital network, represented by a server 166 via the network interface 158.
- the connection to the digital network may be wired or wireless.
- the network interface 158 may be a network port adapted to 10/100/1000 Mbps data traffic, such as an Ethernet port, a modular port being arranged to receive a modular connector, e.g., a RJ45 connector.
- a RJ45 connector port is arranged to receive a network cable, such as a twisted pair cable (e.g., of cat 5, cat 5e or cat 6).
- the I/O means of the network port may be a wireless I/O means using mobile internet communication standards (e.g., 1 G, 2G, 2.5G, 2.75G, 3G, 3.5G, 3.75G, 3.9G, 4G, 5G) or using WiFi.
- mobile internet communication standards e.g., 1 G, 2G, 2.5G, 2.75G, 3G, 3.5G, 3.75G, 3.9G, 4G, 5G
- WiFi Wireless Fidelity
- the camera components i.e. the lens arrangement 146 and the image sensor 150, may be arranged to capture raw images wherein each raw image can be described as light of different wavelengths and originating from different objects and parts of objects. These raw images are then converted from analog to digital format and transferred into the image processing unit 152.
- the digital network camera 142 is a camera arranged to capture visible-light images.
- the image sensor 150 of the digital network camera 142 may be arranged to capture thermal images, or other types of images.
- the object identifier 154 is configured to detect objects of interest in the video sequence captured by the camera 142.
- the object of interest may e.g. be a human for the purposes of the present invention, but it may also be a vehicle, an animal, etc.
- the object identifier 154 is further configured to identify the object of interest in one or more image frames of the video sequence.
- the system further comprises a user interface 168, via which a user may send commands and perform other communication with the circuitry of the camera.
- a single system may comprise several cameras of the disclosed type, all connected to a digital network, preferably connected to the same digital network.
- the user interface 168 may connect to the camera 142 directly, or via the server 166.
- the intelligence may be located separate from the camera, such as on the server 166 implying that the camera basically only serves the purpose of collecting a video stream and forwarding it, with no or limited processing, to another network connected device.
- the system is thus configured to perform the method according to one or any embodiment of the present invention.
- Fig. 6 is a schematic view of a computer program product, comprising a computer-readable storage medium with instructions adapted to carry out the inventive method according to one or any of its embodiments when executed by a device having processing capability.
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Psychiatry (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Social Psychology (AREA)
- Human Computer Interaction (AREA)
- Signal Processing (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Alarm Systems (AREA)
- Image Analysis (AREA)
- Closed-Circuit Television Systems (AREA)
Priority Applications (6)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP16205866.3A EP3340104B1 (de) | 2016-12-21 | 2016-12-21 | Verfahren zur alarmerzeugung in einem videoüberwachungssystem |
TW106141512A TWI749113B (zh) | 2016-12-21 | 2017-11-29 | 在視訊監控系統中產生警示之方法、系統及電腦程式產品 |
JP2017232794A JP7229662B2 (ja) | 2016-12-21 | 2017-12-04 | ビデオ監視システムで警告を発する方法 |
CN201711327621.0A CN108230594B (zh) | 2016-12-21 | 2017-12-13 | 一种用于在视频监控系统中生成警报的方法 |
KR1020170173391A KR102553883B1 (ko) | 2016-12-21 | 2017-12-15 | 비디오 감시 시스템에서 경보들을 발생시키는 방법 |
US15/851,494 US10510234B2 (en) | 2016-12-21 | 2017-12-21 | Method for generating alerts in a video surveillance system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP16205866.3A EP3340104B1 (de) | 2016-12-21 | 2016-12-21 | Verfahren zur alarmerzeugung in einem videoüberwachungssystem |
Publications (3)
Publication Number | Publication Date |
---|---|
EP3340104A1 true EP3340104A1 (de) | 2018-06-27 |
EP3340104C0 EP3340104C0 (de) | 2023-11-29 |
EP3340104B1 EP3340104B1 (de) | 2023-11-29 |
Family
ID=57755012
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
EP16205866.3A Active EP3340104B1 (de) | 2016-12-21 | 2016-12-21 | Verfahren zur alarmerzeugung in einem videoüberwachungssystem |
Country Status (6)
Country | Link |
---|---|
US (1) | US10510234B2 (de) |
EP (1) | EP3340104B1 (de) |
JP (1) | JP7229662B2 (de) |
KR (1) | KR102553883B1 (de) |
CN (1) | CN108230594B (de) |
TW (1) | TWI749113B (de) |
Families Citing this family (19)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109271847B (zh) | 2018-08-01 | 2023-04-07 | 创新先进技术有限公司 | 无人结算场景中异常检测方法、装置及设备 |
EP3629226B1 (de) * | 2018-09-26 | 2020-11-25 | Axis AB | Verfahren zur umwandlung von alarmen |
KR101942808B1 (ko) | 2018-11-13 | 2019-01-29 | 주식회사 인텔리빅스 | 객체 이미지 인식 dcnn 기반 cctv 영상분석장치 |
CN113168606B (zh) * | 2018-11-26 | 2023-09-26 | 埃尔森有限公司 | 用于流程整形的系统和方法 |
KR102127276B1 (ko) | 2018-12-11 | 2020-06-26 | 주식회사 인텔리빅스 | 복수의 고해상도 카메라들을 이용한 파노라마 영상 감시 시스템 및 그 방법 |
KR102194499B1 (ko) | 2019-01-18 | 2020-12-23 | 주식회사 인텔리빅스 | 객체 이미지 인식 dcnn 기반 cctv 영상분석장치 및 그 장치의 구동방법 |
CN109831648A (zh) * | 2019-01-24 | 2019-05-31 | 广州市天河区保安服务公司 | 防盗远程监控方法、装置、设备及存储介质 |
RU196355U1 (ru) * | 2019-06-26 | 2020-02-26 | Леся Николаевна Анищенко | Устройство для автоматизированного распознавания поведения с целью выявления агрессии |
KR102421043B1 (ko) | 2020-02-13 | 2022-07-14 | 주식회사 인텔리빅스 | 영상처리장치 및 그 장치의 구동방법 |
CN111310665A (zh) * | 2020-02-18 | 2020-06-19 | 深圳市商汤科技有限公司 | 违规事件检测方法及装置、电子设备和存储介质 |
KR102201096B1 (ko) | 2020-06-11 | 2021-01-11 | 주식회사 인텔리빅스 | 실시간 cctv 영상분석장치 및 그 장치의 구동방법 |
CN112183252A (zh) * | 2020-09-15 | 2021-01-05 | 珠海格力电器股份有限公司 | 视频的动作识别方法、装置、计算机设备和存储介质 |
US20220083781A1 (en) * | 2020-09-17 | 2022-03-17 | Nec Laboratories America, Inc. | Rule enabled compositional reasoning system |
CN112423021B (zh) * | 2020-11-18 | 2022-12-06 | 北京有竹居网络技术有限公司 | 视频的处理方法、装置、可读介质和电子设备 |
KR102259377B1 (ko) | 2020-12-14 | 2021-05-31 | 김경환 | 인쇄용 롤러 거치장치 |
CN114694197A (zh) * | 2020-12-29 | 2022-07-01 | 中兴通讯股份有限公司 | 行车预警方法、电子设备和计算机可读存储介质 |
WO2023100105A1 (en) * | 2021-12-01 | 2023-06-08 | Ramot At Tel-Aviv University Ltd. | System and method for identifying a person in a video |
CN115394026A (zh) * | 2022-07-15 | 2022-11-25 | 安徽电信规划设计有限责任公司 | 一种基于5g技术的智能监控方法及系统 |
CN115780938A (zh) * | 2022-12-05 | 2023-03-14 | 北京天科合达半导体股份有限公司 | 一种线切割机的切割线监测方法及装置 |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102855462A (zh) | 2011-07-01 | 2013-01-02 | 富士通株式会社 | 确定人体动作周期及识别人体动作的方法和装置 |
CN103106394A (zh) | 2012-12-24 | 2013-05-15 | 厦门大学深圳研究院 | 一种视频监控中的人体行为识别方法 |
US20140341433A1 (en) * | 2004-09-24 | 2014-11-20 | Objectvideo, Inc. | Method for finding paths in video |
Family Cites Families (37)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20050146605A1 (en) | 2000-10-24 | 2005-07-07 | Lipton Alan J. | Video surveillance system employing video primitives |
JP2003087771A (ja) * | 2001-09-07 | 2003-03-20 | Oki Electric Ind Co Ltd | 監視システム及び方法 |
US7058205B2 (en) * | 2001-12-07 | 2006-06-06 | Xerox Corporation | Robust, on-line, view-based appearance models for visual motion analysis and visual tracking |
US7373395B2 (en) * | 2004-02-04 | 2008-05-13 | Perseus Wireless, Inc. | Method and system for providing information to remote clients |
CA2563478A1 (en) * | 2004-04-16 | 2005-10-27 | James A. Aman | Automatic event videoing, tracking and content generation system |
US20060018516A1 (en) * | 2004-07-22 | 2006-01-26 | Masoud Osama T | Monitoring activity using video information |
GB2429515B (en) * | 2005-08-11 | 2008-06-25 | Rolls Royce Plc | Cooling method and apparatus |
US8131012B2 (en) * | 2007-02-08 | 2012-03-06 | Behavioral Recognition Systems, Inc. | Behavioral recognition system |
JP4318724B2 (ja) | 2007-02-14 | 2009-08-26 | パナソニック株式会社 | 監視カメラ及び監視カメラ制御方法 |
US20090036331A1 (en) * | 2007-08-03 | 2009-02-05 | Smith Ian D | Hydraulic fluid compositions |
JP5370927B2 (ja) | 2007-08-21 | 2013-12-18 | 日本電気株式会社 | 行動監視システム、及び、行動監視方法 |
JP2010056720A (ja) * | 2008-08-27 | 2010-03-11 | Panasonic Corp | ネットワークカメラおよびネットワークカメラシステム |
JP4709889B2 (ja) * | 2008-11-20 | 2011-06-29 | パナソニック株式会社 | 移動物体監視装置 |
FR2944629B1 (fr) * | 2009-04-17 | 2017-01-20 | Univ De Tech De Troyes | Systeme et procede de localisation de cible par un reseau de cameras |
MY176067A (en) | 2009-08-24 | 2020-07-24 | Mimos Berhad | System and method to determine suspicious behavior |
US8478048B2 (en) * | 2010-07-08 | 2013-07-02 | International Business Machines Corporation | Optimization of human activity determination from video |
CN102457712A (zh) | 2010-10-28 | 2012-05-16 | 鸿富锦精密工业(深圳)有限公司 | 可疑目标识别及追踪系统及方法 |
TWM406237U (en) * | 2010-12-17 | 2011-06-21 | Utechzone Co Ltd | Intersection traffic pre-warning system |
CN102752574B (zh) * | 2011-04-18 | 2015-01-28 | 中兴通讯股份有限公司 | 一种视频监控系统及方法 |
AU2011201953B2 (en) * | 2011-04-29 | 2013-09-19 | Canon Kabushiki Kaisha | Fault tolerant background modelling |
US9117147B2 (en) * | 2011-04-29 | 2015-08-25 | Siemens Aktiengesellschaft | Marginal space learning for multi-person tracking over mega pixel imagery |
US20160132754A1 (en) * | 2012-05-25 | 2016-05-12 | The Johns Hopkins University | Integrated real-time tracking system for normal and anomaly tracking and the methods therefor |
KR101394270B1 (ko) * | 2012-11-14 | 2014-05-13 | 한국과학기술연구원 | 영상 감지 시스템 및 방법 |
CN102968802A (zh) * | 2012-11-28 | 2013-03-13 | 无锡港湾网络科技有限公司 | 一种基于视频监控的运动目标分析跟踪方法及系统 |
WO2014208575A1 (ja) * | 2013-06-28 | 2014-12-31 | 日本電気株式会社 | 映像監視システム、映像処理装置、映像処理方法および映像処理プログラム |
KR101378071B1 (ko) * | 2013-08-20 | 2014-03-28 | (주)금성보안 | 양식장 및 인공어초 도난 방지 감시 시스템 및 방법 |
KR101484263B1 (ko) * | 2013-10-24 | 2015-01-16 | 주식회사 에스원 | 폭력 감지 시스템 및 폭력 감지 방법 |
JP6411373B2 (ja) * | 2013-12-17 | 2018-10-24 | シャープ株式会社 | 認識データ伝送装置、認識データ記録装置及び認識データ記録方法 |
KR101485022B1 (ko) * | 2014-10-20 | 2015-01-21 | 에코아이넷 주식회사 | 행동 패턴 분석이 가능한 객체 추적 시스템 및 이를 이용한 방법 |
EP3016382B1 (de) * | 2014-10-27 | 2016-11-30 | Axis AB | Überwachungsverfahren und -vorrichtungen |
CN104363414A (zh) * | 2014-10-31 | 2015-02-18 | 无锡悟莘科技有限公司 | 一种智能视频的分析方法 |
IL236752B (en) * | 2015-01-15 | 2019-10-31 | Eran Jedwab | Integrated security method and system |
US9443320B1 (en) * | 2015-05-18 | 2016-09-13 | Xerox Corporation | Multi-object tracking with generic object proposals |
CN104915655A (zh) * | 2015-06-15 | 2015-09-16 | 西安电子科技大学 | 一种多路监控视频的管理方法与设备 |
US9818043B2 (en) * | 2015-06-24 | 2017-11-14 | Microsoft Technology Licensing, Llc | Real-time, model-based object detection and pose estimation |
US20170098172A1 (en) * | 2015-10-06 | 2017-04-06 | Evolv Technologies, Inc. | Training Artificial Intelligence |
US10297026B1 (en) * | 2016-11-29 | 2019-05-21 | Amazon Technologies, Inc. | Tracking of a dynamic container in a video |
-
2016
- 2016-12-21 EP EP16205866.3A patent/EP3340104B1/de active Active
-
2017
- 2017-11-29 TW TW106141512A patent/TWI749113B/zh active
- 2017-12-04 JP JP2017232794A patent/JP7229662B2/ja active Active
- 2017-12-13 CN CN201711327621.0A patent/CN108230594B/zh active Active
- 2017-12-15 KR KR1020170173391A patent/KR102553883B1/ko active IP Right Grant
- 2017-12-21 US US15/851,494 patent/US10510234B2/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20140341433A1 (en) * | 2004-09-24 | 2014-11-20 | Objectvideo, Inc. | Method for finding paths in video |
CN102855462A (zh) | 2011-07-01 | 2013-01-02 | 富士通株式会社 | 确定人体动作周期及识别人体动作的方法和装置 |
CN103106394A (zh) | 2012-12-24 | 2013-05-15 | 厦门大学深圳研究院 | 一种视频监控中的人体行为识别方法 |
Non-Patent Citations (3)
Title |
---|
ALBANESE M ET AL: "A Constrained Probabilistic Petri Net Framework for Human Activity Detection in Video", IEEE TRANSACTIONS ON MULTIMEDIA, IEEE SERVICE CENTER, PISCATAWAY, NJ, US, vol. 10, no. 6, 1 October 2008 (2008-10-01), pages 982 - 996, XP011236996, ISSN: 1520-9210, DOI: 10.1109/TMM.2008.2001369 * |
ALBERTO AMATO ET AL: "Semantic Classification of Human Behaviors in Video Surveillance Systems", 3 October 2011 (2011-10-03), XP055388306, Retrieved from the Internet <URL:http://www.wseas.us/e-library/transactions/computers/2011/54-118.pdf> [retrieved on 20170705] * |
GEORGIA GKIOXARI; ROSS GIRSHICK; JITENDRA MALIK: "Contextual Action Recognition with R*CNN", ARXIV:1505.01197 |
Also Published As
Publication number | Publication date |
---|---|
TWI749113B (zh) | 2021-12-11 |
EP3340104C0 (de) | 2023-11-29 |
JP2018139403A (ja) | 2018-09-06 |
US20180174412A1 (en) | 2018-06-21 |
JP7229662B2 (ja) | 2023-02-28 |
CN108230594B (zh) | 2021-06-29 |
US10510234B2 (en) | 2019-12-17 |
CN108230594A (zh) | 2018-06-29 |
EP3340104B1 (de) | 2023-11-29 |
KR102553883B1 (ko) | 2023-07-11 |
KR20180072561A (ko) | 2018-06-29 |
TW201826141A (zh) | 2018-07-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US10510234B2 (en) | Method for generating alerts in a video surveillance system | |
US10489660B2 (en) | Video processing with object identification | |
AU2017372905B2 (en) | System and method for appearance search | |
CN108053427B (zh) | 一种基于KCF与Kalman的改进型多目标跟踪方法、系统及装置 | |
JP6018674B2 (ja) | 被写体再識別のためのシステム及び方法 | |
TWI759286B (zh) | 用於藉由機器學習訓練物件分類器之系統及方法 | |
EP3321844B1 (de) | Aktionserkennung in einer videosequenz | |
CN110933955B (zh) | 基于对来自摄像机图像的对象的检测的警报事件的改进生成 | |
US10445885B1 (en) | Methods and systems for tracking objects in videos and images using a cost matrix | |
WO2021223665A1 (zh) | 一种减少无效告警的方法及装置 | |
KR20200058260A (ko) | 객체 이미지 인식 dcnn 기반 cctv 영상분석장치 및 그 장치의 구동방법 | |
US10140554B2 (en) | Video processing | |
WO2016201683A1 (en) | Cloud platform with multi camera synchronization | |
KR20210062256A (ko) | 행동 시퀀스 기반으로 이상행동 여부를 판단하는 방법, 프로그램 및 시스템 | |
CN113920585A (zh) | 行为识别方法及装置、设备和存储介质 | |
KR102111162B1 (ko) | 블랙 박스와 연동된 다채널 카메라 홈 모니터링 시스템 및 방법 | |
Rathour et al. | KlugOculus: A Vision‐Based Intelligent Architecture for Security System | |
KR20220118857A (ko) | 영상 분석 서버의 분석 데이터를 이용한 영상 검색 시스템 | |
KR20230012171A (ko) | 딥러닝 기반의 tod 객체 탐지 후 식별 시스템 및 방법 | |
CN118015770A (zh) | 电子围栏命中方法、终端设备及计算机可读存储介质 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PUAI | Public reference made under article 153(3) epc to a published international application that has entered the european phase |
Free format text: ORIGINAL CODE: 0009012 |
|
STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: REQUEST FOR EXAMINATION WAS MADE |
|
17P | Request for examination filed |
Effective date: 20171018 |
|
AK | Designated contracting states |
Kind code of ref document: A1 Designated state(s): AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR |
|
STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: EXAMINATION IS IN PROGRESS |
|
17Q | First examination report despatched |
Effective date: 20200714 |
|
STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: EXAMINATION IS IN PROGRESS |
|
RAP1 | Party data changed (applicant data changed or rights of an application transferred) |
Owner name: AXIS AB |
|
STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: EXAMINATION IS IN PROGRESS |
|
REG | Reference to a national code |
Ref document number: 602016084371 Country of ref document: DE Ref country code: DE Ref legal event code: R079 Free format text: PREVIOUS MAIN CLASS: G06K0009000000 Ipc: G06V0020520000 |
|
RIC1 | Information provided on ipc code assigned before grant |
Ipc: G06V 40/20 20220101ALI20230719BHEP Ipc: G06V 20/52 20220101AFI20230719BHEP |
|
GRAP | Despatch of communication of intention to grant a patent |
Free format text: ORIGINAL CODE: EPIDOSNIGR1 |
|
STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: GRANT OF PATENT IS INTENDED |
|
INTG | Intention to grant announced |
Effective date: 20230829 |
|
GRAS | Grant fee paid |
Free format text: ORIGINAL CODE: EPIDOSNIGR3 |
|
GRAA | (expected) grant |
Free format text: ORIGINAL CODE: 0009210 |
|
STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: THE PATENT HAS BEEN GRANTED |
|
AK | Designated contracting states |
Kind code of ref document: B1 Designated state(s): AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR |
|
REG | Reference to a national code |
Ref country code: GB Ref legal event code: FG4D |
|
REG | Reference to a national code |
Ref country code: CH Ref legal event code: EP |
|
REG | Reference to a national code |
Ref country code: DE Ref legal event code: R096 Ref document number: 602016084371 Country of ref document: DE |
|
REG | Reference to a national code |
Ref country code: IE Ref legal event code: FG4D |
|
U01 | Request for unitary effect filed |
Effective date: 20231211 |
|
U07 | Unitary effect registered |
Designated state(s): AT BE BG DE DK EE FI FR IT LT LU LV MT NL PT SE SI Effective date: 20231215 |
|
PGFP | Annual fee paid to national office [announced via postgrant information from national office to epo] |
Ref country code: GB Payment date: 20231219 Year of fee payment: 8 |
|
U20 | Renewal fee paid [unitary effect] |
Year of fee payment: 8 Effective date: 20240111 |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: GR Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20240301 |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: IS Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20240329 |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: ES Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20231129 |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: IS Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20240329 Ref country code: GR Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20240301 Ref country code: ES Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20231129 |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: RS Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20231129 Ref country code: PL Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20231129 Ref country code: NO Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20240229 Ref country code: HR Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20231129 |